Navigation – Plan du site

AccueilNuméros11(1)Comptes rendusZygmunt Frajzyngier, Locative pre...

Comptes rendus

Zygmunt Frajzyngier, Locative predications in Chadic languages: Implications for semantic analysis

Joseph Lovestrand
Référence(s) :

Zygmunt Frajzyngier, Locative predications in Chadic languages: Implications for semantic analysis, Oxford, Oxford University Press, 2024, doi: 10.1093/oso/9780198896210.001.0001.

Texte intégral

1Zygmunt Frajzyngier, Professor Emeritus at the University of Colorado, is a prolific author continually seeking to expand our scholarly understanding of language, in general, and Chadic languages, in particular. This most recent book discusses eight Chadic languages. Each language is treated separately in its own chapter, presenting data from previous publications by the same author with some changes to the analyses. These eight chapters are preceded by two short chapters of preliminaries and followed by two short chapters of general summary and discussion.

2The book opens with a stated aim of discussing “semantic structures” which “should be the main object of semantic description” (p. 1). These “semantic structures” are not formal representations of the meaning of individual words or expressions (as in Jackendoff 1990), but rather the totality of “functional domains” expressed across the grammar of a particular language (p. 10, 313). The focus of this book is limited to the “locative domain”, and so each of the eight main chapters presents a range of overlapping topics related to the morphosyntax of location and motion in that language.

  • 1 Example 11 on p. 88 shows some unexplained variation in this pattern. The same verb and locative co (...)

3One of the morphosyntactic patterns of interest to Frajzyngier is the (adpositional) marking of locative complements, that is, noun phrases or adverbs with a semantic role of location, goal or source. This theme expands on Frajzyngier (2022a), a paper which discusses five of the eight languages in this monograph. Cross-linguistically, many languages have a class of nouns which can appear as unmarked locative complements, at least in certain contexts. For example, Frajzyngier presents data on Pero, where the preposition precedes all types of locative complements, whether stative location, goal or source. However, for toponyms and a select number of common nouns (e.g. pídì ‘place’, mínà ‘house’, láɗi ‘church’), the preposition is omitted when the noun phrase functions as a goal (p. 93).1 The preposition still appears if the locative complement is a location or source (p. 96-97).

4This pattern fits into previous observations, most recently discussed under the label of differential place marking, that unmarked locatives tend to be toponyms in the semantic role of goal (Haspelmath 2019; Stolz et al. 2014: 287). The Chadic data Frajzyngier presents include some noteworthy variants of differential place marking. In Mina, the widespread locative preposition n is not used with human referents; instead, the form r is used. As Skilton & Obert (2022) point out, differential place marking in cases like this is not a choice between less and more phonological material, but an alternation between overt markers of equal size. Another intriguing pattern is found in Frajzyngier’s description of Mina: a preposition á (opaquely labeled “locative predicator”) appears before a locative complement if the predicate is from a class of “non-locative verbs”, and is otherwise absent. This appears similar to differential place marking determined by verb class in the Amazonian language Ticuna (Skilton & Obert 2022: 16). On analogy with other types of differential marking, Skilton & Obert (2022: 4–5) speculate that differential place marking could be (partly) conditioned by tense-aspect. Frajzyngier’s description of Lele includes a preposition , which is apparently used to mark at least some goals “when they code past events” (p. 126).

5Despite the parallels between recent work on differential place marking and Frajzyngier’s data from Chadic languages, Frajzyngier does not engage with the relevant literature. Instead of a fine-grained analysis of variation in differential place marking, Frajzyngier proposes a dichotomous typology between languages which have a “locative domain” and those that do not. The presence or absence of the “locative domain” is said to be a “fundamental distinction” that has “not been observed or described in other languages” (p. 2) and yet it has “far-reaching consequences for semantic theory and the methodology of semantic research” (p. 4).

  • 2 Elsewhere, a “functional domain” is said to be composed of “a finite number of functions encoded wi (...)
  • 3 Most of the same locative functions found in other Chadic languages are also present in the morphos (...)
  • 4 Stolz et al. (2014: 37) discuss the case of English home as the sole unmarked locative and conclude (...)

6As a summary of the various descriptions and definitions of “locative domains” throughout the book, Frajzyngier concludes that: “There exist different types of locative domains. Their common feature is the existence of verbs, nouns, or both classes of lexical items that are inherently locative and others are not” (p. 314).2 These “inherently locative” lexical items include, for example, noun classes that can be unmarked as locatives. In other words, diagnosing a “locative domain” appears to be the equivalent of identifying a grammatical system that includes a class of unmarked locatives. Of the eight Chadic languages discussed, one language, Gidar, is said not to have evidence of a “locative domain”, i.e., no unmarked locatives.3 However, this claim is contradicted by one noun, íŋgílì ‘home’, which is unmarked when in the role of goal (p. 298).4 Therefore, none of the Chadic languages discussed in this book demonstrate the “fundamental distinction” between languages with a “locative domain” and those without.

7Despite their absence from this book, there do exist languages without unmarked locatives, so, for the sake of argument, we can move on to consider the supposed implications of the proposed typological dichotomy. The distinction between languages with and without a “locative domain” is meant to “provide evidence for the directionality of a cause-and-effect relationship with respect to the question whether the properties of verbs determine the form of the locative predication or whether the functions encoded in the locative complements affect the syntactic properties of verbs” concluding that “the latter is the case” (p. 13). Frajzyngier rightly recognizes that this position “runs against a frequent assumption that the choice of verbs determines the forms of clauses” (p. 188).

8For example, in the description of Lele, Frajzyngier states that the “locative domain coded in the semantic structure of a language determines the properties of the lexical categories ‘verb’ and ‘noun’, in that some verbs and some nouns are inherently locative, and others are not” (p. 138).

  • 5 To confirm how a “locative domain” is defined in this context, we read on the previous page: “The e (...)

9If the “locative domain” is defined as a grammatical system that has unmarked locatives (i.e., “inherently locative” nouns),5 then this claim is circular: the existence of a class of unmarked locatives determines that there is a class of unmarked locatives. Elsewhere, Frajzyngier writes, “[o]ne of the outcomes of this distinction [between languages with and without a locative domain] is the presence of inherently locative verbs and inherently locative nouns, which form distinctive syntactic classes within the lexicon” (p. 2). Again, the reasoning is circular. Having defined a distinction based on some morphosyntactic feature, in what sense is the presence of that same feature the “outcome” of the distinction the linguist made?

  • 6 Frajzyngier also claims that je is historically developed from an erstwhile verb, but this claim is (...)

10Overall, the conceptual framework Frajzyngier employs is not an improvement on current comparative approaches to this topic. Furthermore, there are a number of issues with the morphosyntactic analyses in this book. For example, in the description of Lele, a morpheme -je is alternately described as a ventive suffix or as a verb meaning ‘come’. Elsewhere, Frajzyngier and others convincingly demonstrate that -je is a suffix, not a verb, so the reason for this variation in the description is unclear (Frajzyngier 2001: 202; Simons Cope 1982: 228; Weibegué & Palayer 1982: 21).6 In the description of Mina, there is mention of a verbal suffix -á “which indicates that the verb has a goal. Such a goal could be an object pronoun, another verb in the clause, or an inferable place” (p. 24). How this description maps onto the distribution of this morpheme is unclear, and Frajzyngier (2022b: 461) previously admitted that “[t]he function of the verbal suffix -a in Mina, glossed as ‘goal’, indeed merits a revised analysis.”

11The most difficult morphosyntactic analysis to follow as a reader is that of Hdi. Frajzyngier is rather ambivalent about a putative morpheme a found at the end of verbal extensions (suffixes), which expresses “directionality to a goal, but the goal itself is unspecified” (p. 141). However, this morpheme is not parsed in the interlinear examples. In the two lists of verbal extensions given, one presents six suffixes composed of a single consonant while the other list adds a vowel a to three of the six suffixes (p. 142, 146). In addition to morpheme a, there is also an “allative” suffix -. This suffix is said to indicate “directionality away from the inferable place” (p. 147) but also “movement toward a goal” (p. 148). The examples given do not provide compelling evidence for this analysis, and some explanation is needed for how this suffix can co-occur with the ventive suffix -gá (p. 146, 163).

12Frajzyngier’s previous work has been critiqued elsewhere for internally inconsistent analyses, unstructured argumentation, circular definitions, and insufficient engagement with relevant literature (Cysouw 2001; Elders 2000; Hellwig 2003; 2022; Jaggar 1994; Löhr 2007; Miestamo 2006; Wolff 2011). This recent publication is no exception to that pattern. Those interested in the grammars of the particular Chadic languages discussed may benefit from comparing the analyses presented in this book with former publications, but this monograph does not represent a significant advancement in the description and comparison of the grammars of Chadic languages. Despite the ambitious and laudable goal of taking a holistic approach to comparing the distribution of meaning across grammatical systems, this work falls short of its stated aims and is not a reliable representation of the state of the art in comparative linguistics. Given the recent steep decline in the number of scholars actively interested in studying Chadic languages, it is especially unfortunate that this prominent publication by a widely recognized scholar cannot be recommended as essential reading.

Haut de page

Bibliographie

Cysouw, Michael. 2001. [Review] A grammar of Lele by Zygmunt Frajzyngier (2001). Linguistlist 12(3099).

Elders, Stefan. 2000. [Review] Grammaticalization of the complex sentence. Lingua 110(4). 305–313. https://doi.org/10.1016/S0024-3841(99)00038-8).

Frajzyngier, Zygmunt. 2001. A grammar of Lele. Stanford, CA: CSLI Publications.

Frajzyngier, Zygmunt. 2022a. Toward a functional typology of adpositions: Theoretical implications. Folia Linguistica. 56(1). 123–151. https://doi.org/10.1515/flin-2021-2006.

Frajzyngier, Zygmunt. 2022b. Meaning encoded in the grammatical system: A rejoinder. Folia Linguistica 56(2). 461–467. https://doi.org/10.1515/flin-2022-2023.

Haspelmath, Martin. 2019. Differential place marking and differential object marking. STUF - Language Typology and Universals 72(3). 313–334. https://doi.org/10.1515/stuf-2019-0013.

Hellwig, Birgit. 2003. [Review] A grammar of Hdi. Journal of African Languages and Linguistics 24(2). 187–190.

Hellwig, Birgit. 2022. Adpositions and their distribution: A reply to Zygmunt Frajzyngier’s “Toward a functional typology of adpositions: theoretical implications.” Folia Linguistica 56(2). 453–459. https://doi.org/10.1515/flin-2022-2022.

Jackendoff, Ray. 1990. Semantic structures. Cambridge, MA: MIT Press.

Jaggar, Philip J. 1994. [Review] Zygmunt Frajzyngier: A dictionary of Mupun. Bulletin of SOAS 57(3). 102. https://doi.org/10.1017/S0041977X0000937X.

Löhr, Doris. 2007. [Review] A grammar of Mina. Journal of African Languages and Linguistics 28(2). 236–238.

Miestamo, Matti. 2006. [Review] Zygmunt Frajzyngier and Erin Shay (2003). Explaining language structure through system interaction. Studies in Language 30(3). 617–626. https://doi.org/10.1075/sl.30.3.08mie.

Simons Cope, Pamela. 1982. Nè...ba marking in Lele: A cleft construction. Studies in African Linguistics 13. 217–229.

Skilton, Amalia Horan & Karolin Obert. 2022. Differential place marking beyond place names: Evidence from two Amazonian languages. Glossa 7(1). https://doi.org/10.16995/glossa.6371.

Stolz, Thomas, Sander Lestrade & Christel Stolz. 2014. The crosslinguistics of zero-marking of spatial relations. Berlin: De Gruyter Mouton. https://doi.org/10.1524/9783050065304.

Weibegué, Christophe & Pierre Palayer. 1982. Lexique lele-français. Sarh: Centre d’Etudes Linguistiques, Collège Charles Lwanga.

Wolff, H. Ekkehard. 2011. Language variation, theoretical preoccupations and the Lamang-Hdi language continuum in Central Chadic. Review of Zygmunt Frajzyngier with Erin Shay: A grammar of Hdi. Afrika und Übersee 90(2008/2009). 21.

Haut de page

Notes

1 Example 11 on p. 88 shows some unexplained variation in this pattern. The same verb and locative complement combination are repeated, once with the preposition ti and once without.

2 Elsewhere, a “functional domain” is said to be composed of “a finite number of functions encoded within the domain” which all “share the defining semantic feature of the domain” (p. 15). It is not immediately apparent why differential place marking is employed as a litmus test for the presence of the “locative domain”.

3 Most of the same locative functions found in other Chadic languages are also present in the morphosyntax of Gidar. These functions are described as “locative expressions” rather than as a “locative domain”.

4 Stolz et al. (2014: 37) discuss the case of English home as the sole unmarked locative and conclude that this should be considered a prototypical instance of unmarked locatives, rather than an idiosyncratic exception. Even if Gidar were to be considered a language without unmarked locatives, note that there is differential place marking of human locatives which are preceded by an additional morpheme ‘hand’ or ‘side’ (p. 283, 291).

5 To confirm how a “locative domain” is defined in this context, we read on the previous page: “The existence of the locative domain is evidenced by the distinction between inherently locative lexical items, toponyms, and locative adverbs and non-locative nouns” (p. 137).

6 Frajzyngier also claims that je is historically developed from an erstwhile verb, but this claim is not substantiated.

Haut de page

Pour citer cet article

Référence électronique

Joseph Lovestrand, « Zygmunt Frajzyngier, Locative predications in Chadic languages: Implications for semantic analysis »Linguistique et langues africaines [En ligne], 11(1) | 2025, mis en ligne le 30 juin 2025, consulté le 08 février 2026. URL : http://journals.openedition.org/lla/14050 ; DOI : https://doi.org/10.4000/15e3z

Haut de page

Auteur

Joseph Lovestrand

University of Essex

Articles du même auteur

Haut de page

Droits d’auteur

CC-BY-SA-4.0

Le texte seul est utilisable sous licence CC BY-SA 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont susceptibles d’être soumis à des autorisations d’usage spécifiques.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search